Skip to content
For space & earth-observation companies

Your imagery operations team,
made of software.

It quotes the order, checks whether you already own the answer before you pay to collect it again, chases every delivery through the pipeline, and reports what changed — with the scene that proves it attached.

Where this actually stands

The platform underneath is in production today, running real businesses. The space ontology and the connectors are not — we build those with the first design partner, against a real fleet and real data rather than a demo.

3
run today, unchanged
2
are a schema pack on an existing engine
4
are genuinely new — the domain work

We'd rather lose you here than in the demo. If you want the full picture of what exists, the feature list is public.

Who this is for

A specific company, not a category

Earth-observation and analytics companies selling insight built on imagery, and satellite operators running constellations of 10–200.

The problem

Your largest variable cost is collection you may not have needed, and nobody has time to check the archive first.

What you do instead today

An analyst orders new imagery because checking what you already hold is tedious, and an ops person refreshes a dashboard to see whether it arrived.

An imagery analyst or mission ops coordinator costs $90k–$160k. That's the line this is priced against — not a per-seat software budget.

The roster

9 employees, each with one job

Every one produces an artifact you can check, and stops at a gate you set.

Order Desk

Running today

Handles inbound requests for imagery or analysis end to end.

What you get
A quoted, feasibility-checked order with an expected delivery window and cost, ready for approval.
Where it stops
Anything above a spend threshold or committing to a delivery SLA.

Archive vs. Tasking

Built with you

Answers the question the cheapest defensible way.

What you get
“The archive answers this at 94% confidence for $0; new tasking costs $1,850 and adds 4%” — with the evidence for both paths.
Where it stops
Spend authorisation on any new collection.

Delivery Pipeline Watch

Adapts from what's built

Makes sure every ordered thing actually arrives, processed correctly, on time.

What you get
Proactive notification on capture and delivery — and on a failure, the reprocess or re-task already initiated.
Where it stops
Re-tasking spend and any customer-facing SLA concession.

Monitoring Analyst

Built with you

Reviews every new scene over your customers' assets and reports only what matters.

What you get
A detection with a confidence score, the proving image with its acquisition metadata, and a recommended action.
Where it stops
Low-confidence detections route to human review automatically.

Ground Segment Coordinator

Built with you

Books downlink capacity, schedules passes, and reconciles the invoices.

What you get
An optimised schedule and a reconciled invoice with the discrepancies flagged.
Where it stops
Spend commitments.

Mission Anomaly Watch

Built with you

Monitors spacecraft telemetry, catches divergence, and drafts the anomaly report with ranked responses.

What you get
An anomaly record with the divergence evidence, similar historical cases, and two or three ranked recommended responses.
Where it stops
Every command, without exception. It recommends; it never executes.

Regulatory & Licensing

Adapts from what's built

Tracks filing deadlines, spectrum coordination, debris compliance, and export requirements.

What you get
Drafted filings with the supporting evidence already assembled.
Where it stops
Everything. Legal signs filings.

Customer Reporting

Running today

Produces the monthly delivery, coverage, and SLA report per customer.

What you get
A formatted report where every number links to the captures underneath it.
Where it stops
Account manager review before it sends.

Margin & Capture Economics

Running today

Computes true cost per capture, per delivered product, per customer.

What you get
Unit economics that drill down to individual captures.
Where it stops
None — it only reads.
What it runs on

The part that already exists

These six subsystems are in production. A vertical changes the objects in the graph and the tools the agents hold — not the engine.

The ontology

A typed graph of your business with row-level tenant isolation, PII classes masked from agents by default, and provenance on every fact.

The action engine

Every state change resolves, validates against a schema, passes an approval gate, checks a precondition, snapshots before and after, and writes exactly one audit row — including the denials.

Governance

A role × risk-tier matrix enforced in the engine rather than the UI, plus per-tenant and global kill switches.

The agent runtime

140 skills, 44 tools, crews that finish multi-step jobs, and delegation between agents with verification.

The front office

Voice, SMS, email, and web — answered 24/7, in over 100 languages, booking into a real calendar.

Cloud compute

The agent's own sandboxed computer for work that outlives a request, metered and capped per plan.

The hard part

What breaks when agents do real work

These are the failures that only appear in production — and the reason a fleet or a constellation is a serious place to run this rather than a demo. Every one is shipped and tested, and none of it is vertical-specific.

Delegation that can't escalate

One agent hands work to another with a strict subset of its own authority — never more. The chain is depth-limited, cycle-free, and draws on one shared budget, so an agent can't delegate its way around a spend cap.

Cost per completed job

Every model call, tool call, and sandbox minute in a chain rolls up to the job a person actually asked for. The figure includes the attempts that failed, because a system that fails half the time at $0.40 a win really costs $0.80.

Outcome evals, not prompt evals

Success is asserted against the records, not the transcript. State is captured before and after, so the agent gets no credit for a booking that already existed — and a job that hit its goal but also messaged the wrong people is scored as a failure.

Memory that can be wrong

Contradictions are caught at write time. What a customer stated outranks what an agent inferred, regardless of which arrived later. Facts decay at a rate set by how fast that kind of fact actually changes, and corrections supersede rather than delete — so you can still reconstruct what was believed on the day a decision was made.

One customer, every channel

The same person across a phone, an email, and a web form resolves to one record. Deliberately cautious: a matching phone and name merges, a matching address and surname does not, because that's a household — and a wrong merge is a privacy incident, not a duplicate.

Lifecycle and offboarding

Dormant, unowned, and over-scoped agents surface without anyone maintaining a list. Retiring one revokes its access immediately, forces its in-flight work to be reassigned, and keeps the full record — an agent with live tasks and no successor cannot be retired at all.

Undo, described honestly

Every connector action declares what reverses it. Most things can only be offset, not undone: a refund is a second transaction the customer can see, and a sent email cannot be recalled at all. You're told which is which before you approve, not after.

Portability

Your agents, ontology, and memory export as a documented open format — including a statement of what it can't contain and why. Credentials never travel; they're the provider's, scoped to us, and exporting them would be a security failure rather than a feature.

The consequence gradient is why this matters more here than in a salon. A wrong invoice is embarrassing; a bad policy release reaches a machine, and a wrong command reaches something you cannot service. Anything at that end is built as a recommender that physically cannot execute.

The model

What space looks like as a graph

SatelliteSensorPassGround StationCaptureObservationProcessing RunOrderDeliveryAOI (area of interest)Asset (the customer's physical thing)DetectionLicenseFiling

Why every fact carries a time

Every observation has a capture time, a downlink time, a processing time, and an ingest time — and a reprocessing can change a conclusion you already sent. “What did we conclude, from which data, using which model version, on which date?” is the only defensible basis for a claim about a physical place.

Build on it

What you'd be building against

The ontology API, the MCP endpoint, and the typed SDK are live and in production. Access is granted directly — there's no public signup for these verticals, so you talk to us and we issue the credentials.

Connect your provider catalogs

Ontology API + typed SDK

Point it at the archives and tasking APIs you already buy from. Every order, capture, and delivery lands in one graph instead of four vendor portals.

Automate the order-to-delivery loop

Actions API + approval gate

Quote, feasibility-check, order, and notify — with a spend gate on anything that commits to new collection.

Expose tasking to your own customers

Scoped API keys

Let them order through your product while the feasibility and pricing logic runs on your side, scoped so one customer can never see another's orders.

Run your detection models as an agent

Agent runtime + tool registry

Bring your own change-detection or classification model, and let the agent handle the scheduling, the confidence routing, and the evidence trail around it.

Ask your operations a question

MCP

“Which orders are late and which provider caused it?” answered from live data, in Slack or in your own console.

How you get access

Developer credentials are issued by us, not signed up for. Tell us what you're building and we'll scope the keys to it — read-only if that's all you need, action-invoking if you're automating something. It's a conversation, not a form, because at this stage every integration teaches us something about the space model.

Request developer access

What we do rather than you

Defining brand-new object types — a Robot, a Satellite, a Pass— runs through us. Type changes reshape a tenant's whole graph, so they go through an internal admin rather than a public endpoint. You query, act, and build agents against the model; the space schema itself is what the design partnership adds.

Straight answers

The questions you're about to ask

Is this available today?+

The platform is — ontology, approval gates, audit trail, agent runtime, and the order-desk front office all run in production. The space schema pack and the provider connectors are not built. We build those with a design partner, against real orders and real deliveries.

Which of these do you build first?+

Order Desk, Archive vs. Tasking, and Delivery Pipeline Watch. They reuse the most of what already exists, and Archive vs. Tasking pays for the engagement on its own by attacking the largest variable cost in the business.

Would an agent ever command a spacecraft?+

No. Mission Anomaly Watch is deliberately built as a recommender with no execution path — the consequence is irreversible and the asset can't be serviced. It surfaces the divergence, finds comparable historical cases, ranks the responses, and stops. A human commands.

We're an analytics company, not an operator. Does this still apply?+

That's the customer we'd start with. The commercial cycle is faster, the budgets are nearer to hand, and Order Desk plus Archive vs. Tasking is immediate margin rather than a procurement exercise.

What does a design partnership involve?+

Access to your real order flow and one operational job you'd otherwise hire for. We build the schema pack and the first three agents against your data. You get them at a price reflecting that you're the reason they work.

We're looking for one space design partner.

Not ten. The first one shapes the schema pack and gets the agents built against their own data — at a price that reflects being the reason it works.

Start the conversation

Most useful if you're already delivering imagery to paying customers.